面向智能家居视频监控的嵌入式Agent的设计与实现

发布时间：2018-02-20 04:36

本文关键词： 智能家居嵌入式技术 MAS系统嵌入式Agent 物联网技术　出处：《北方民族大学》2016年硕士论文　论文类型：学位论文

【摘要】：信息全球化和家庭网络的兴起极大地改变了人们的家居生活方式,智能家居的概念应运而生。而嵌入式技术、网络技术以及人工智能技术等各学科的交叉应用,为智能家居的发展提供了新的思路和方法。在未来的智能家居系统中,如何使分散的、不同种类、不同结构的家电设备联系起来,通过协调与协作共同完成一些复杂的任务是智能家居系统需要解决的主要问题。而多Agent体系提供了一个很好的解决方案。并且MAS体系对于新家电的加入、退出提供比较宽松和灵活的约束,这对处于动态的家居环境具有重要意义。本课题以智能家居这个小型分布式控制系统为研究对象,主要进行了两方面的研究工作。第一利用MAS理论构建了智能家居系统体系,该体系包括三层,最上层为战略层Agent,主要负责多Agent协调与用户的交互,是系统的最高决策层;中间层为管理层,是整个系统的桥梁,接收经过上层Agent粗粒度划分的控制子命令,并要将子命令进一步细分为元命令,分配给下层Agent。而当自身无法完成任务命令时,要与同级Agent连接,请求协作;最下层为控制层,主要实现设备控制,并将执行情况反馈给其上层Agent。第二根据这个框架,本文在软硬件设计与系统实现上设计并实现了一个智能家居系统。其中,最上层Agent在PC机上实现,中间层Agent在mini2440上实现。中间层Agent与最下层Agent的通信用ZigBee无线通信协议实现。最下层的设备控制以智能家居系统中的重要组成部分视频监控为对象。最后对系统做了集成测试,测试该系统的稳定性、智能性。本文旨在为多Agent理论研究与实际应用相结合做一个案例,能为往后智能家居系统设计与研究的方向做一个探索。
[Abstract]:The globalization of information and the rise of home network have greatly changed people's way of living at home, and the concept of smart home has come into being. However, the intersecting application of embedded technology, network technology and artificial intelligence technology, etc. It provides new ideas and methods for the development of smart home. In the future smart home system, how to make the decentralized, different types, different structure of household appliances connected together, To accomplish some complex tasks through coordination and cooperation is the main problem that smart home system needs to solve. The multi-#en0# system provides a good solution, and the MAS system adds to the new household appliances. Exit provides relatively loose and flexible constraints, which is of great significance to the dynamic home environment. In this paper, smart home, a small distributed control system, is taken as the research object. Firstly, the intelligent home system is constructed by using MAS theory. The system consists of three layers, the top layer is the strategic layer, which is mainly responsible for the interaction between multiple Agent and users, and is the highest decision-making level of the system. The middle layer is the management layer, which is the bridge of the whole system. It receives the control subcommands that are partitioned by the coarse granularity of the upper Agent, and further subdivides the sub-commands into meta commands and assigns them to the lower level agents. To connect with the peer Agent, request collaboration; the lowest layer is the control layer, which mainly implements device control, and feedback the performance to its upper agent. Second, according to this framework, In this paper, a smart home system is designed and implemented in hardware and software design and system implementation. The intermediate layer Agent is implemented on mini2440. The communication protocol of the middle layer Agent and the lowest layer Agent is implemented. The device control of the lowest layer takes video surveillance, an important part of the smart home system, as the object. Finally, the integration test is done to the system. The purpose of this paper is to make a case for the combination of multiple Agent theory research and practical application, and to explore the direction of the design and research of smart home system in the future.
【学位授予单位】：北方民族大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TU855;TN948.6

【相似文献】